Developing audio engines with live coding capabilities
I've always had an anxiety implementing live coding systems. I'm never convinced I did it right and that it's only barely working.
I added hot swapping to Sporth years ago, which made live coding possible, but it required moving tons of code around, which made me suspicious.
My current system does live coding in a similar way, and it's behaved pretty well for me for years. But, more recently, I also now have the ability to crossfade between states. A bit more fancy. It seems to work, it sounds awesome, but the low level anxiety is still there. Like wondering if you left your stove on.
I've been recently revisiting the system after months away. The brain has a way of smoothing over fine details, and with it, some of the specific concerns I had about it in the first place. As I start working with it more, I find myself slowly building up trust.